Industrial wastewater treatment control by a minimax principle over weakly measured pollution

Abstract

In this paper, a problem of rationally treating industrial wastewater is considered. The treatment is conditioned with that a one or more plants should execute this job along a river but a schedule of treating the polluted water is not correlated with the schedule of manufacturing. So, river water is treated separately, without considering manufacturing processes. Another condition is that they deal with scattered and dependent-on-location measurements of water pollution. Thus, it is hard to determine exactly for each plant a volume of water that should be necessarily treated for ensuring the water renewal for the whole river. Because of multiple, weakly correlated measurements, industrial water treatment should be controlled both for dealing with water pollution and prevention of detrimental effects. A robust approach is proposed to industrial wastewater treatment control with using a minimax principle, which fits long, deep, and wide rivers. Pollution at a shallow river can be analyzed as well only if its contaminations are registered in wide ranges. The control realized by a minimax model and algorithm is meant by effectively assigning treatment jobs for wastewater treatment plants. In three special cases, the model gives hints at plausible inaccurate measurements or overestimations.